A NON-INVASIVE CLINICAL DECISION SUPPORT AND EARLY WARNING SYSTEM FOR CORTICAL SPREADING DEPRESSION AND ITS METHODOLOGY

Abstract

The invention allows for the treatment of Cortical Spreading Depression (CSD) crises without the need for live subjects. Simulating synthetic surface EEG data for training deep learning-based models. to produce; - - - - - receiving a virtual external input signal (p) and a pyramidal cell output population macroscopic dynamics between excitatory and inhibitory interneuron populations A Jansen & Rit system that calculates (gas-brake dynamics) using differential equations. Neural Mass Modulus (1.1), The cell voltage potentials obtained from the module in question are defined in advance. Neuronal firing between maximum and minimum biological threshold values a sigmoid function transformation unit (1.2) that converts to frequency, the brain's electrical background noise and the Wiener process in question by including a random input to the differential equations, the system equations An Euler-Maruyama module (1.3) that prevents divergence and stabilizes the system. Sodium-potassium (Na+ / K+) ion pumps that occurred during the CSD crisis microscopic failure at the cellular level and synaptic gain parameters (A and B) deactivation cycles, a macroscopic recovery parameter by reducing the processor (CPU) load, that is, within a defined time by reducing it to a mathematical coefficient that decreases and increases within a range a fabric that dynamically changes the parameters of differential equations Recovery time constant (?) manipulation module (1.4) and The ignition produced by the sigmoid function conversion unit (1.2) in question. receiving frequency data, different spatial representations of virtual electrode coordinates By solving equations on (spatial) nodes, low-pass cranial bone simulating the filter effect with mathematical attenuation coefficients and the skull converting underlying CSD signal components into synthetic surface EEG signal data a deep learning-based non-invasive clinical containing a spatial network processing unit (1.5) decision support and early warning simulation system (1) and its working method It is related.

Images